Randomisation, Allocation Concealment, ITT Analysis, And Pragmatic And Explanatory Trials Are Essential In Essay.

1: Define randomisation and outline the three reasons why randomisation is an important design element in trials assessing interventions

2: Define allocation concealment and discuss how it differs from blinding.

3: Define and discuss intention to treat analysis and why is it the preferred mode of analysis in trials compared to other types of analysis.

4: Define, compare and contrast pragmatic and explanatory trials. Explain why a pragmatic trial might be more useful than an explanatory trial.

Importance of Randomisation in Trials Assessing Interventions

In Randomised Control Trial, participants are randomly allocated either to the control group (group receiving standard treatment) or to the treatment group under investigation. This is called as the process of randomisation as it reduces the bias in selection and allows the researcher to compare the results of the treatment with the no treatment group while keeping other variables constant (Paludan-Müller et al., 2016).

It is useful in assessing intervention as it provides information on the adverse effects of the treatment while eliminating the bias in the treatment assignment. Due to reduce of spurious casualty the results from the trials are reliable form of scientific evidence (Schulz, 1995). Randomisation facilitates blinding the “identity of treatment” from the participants, investigators and the assessors. Therefore, concealing the nature of intervention ensures fairness in results as it limits the possibility of the contamination (Day & Altman, 2000). Further, randomisation permits using probability theory. It expresses the likelihood that the “differences in the outcome” between the treated groups merely indicates “chance”. Hence, randomisation is an important design element in trials assessing interventions (Gore, 1981).

The process of protecting the randomisation so that the identity of the treatment to be allocated is not known to the patient before entering the study is known as allocation concealment. In short the treatment allocation system is set up in a manner that does not reveal the person enrolling the participants as to what treatment will be given to them.  If the investigators are aware of the adverse effects of the intervention it may reveal the allocation to the assessors which may in turn induce bias (Altman & Schulz, 2001). However, the concept of allocation concealment is different from that of blinding.  

In blinding the participants in trial and their family members, caregivers, adductors and investigators are unaware of the treatment identity. In contrast, in the allocation concealment only the trial participants and the person enrolling them are kept unaware Blinding helps to prevent the clinicians from treating the patients (consciously or subconsciously) based on the allocation. It also prevents the observer bias. Blinding does not allow the  assessors to introduce bias when assessing the subjective outcome. Even the patients cannot introduce bias during flow up (Sedgwick, 2013). The primary difference between the allocation concealment and blinding is that former ensures that the treatment identity is not revealed before the participant enters into study. The later ensures that the patient or the physician does not know the treatment allocation after the enrolment into the study (Ruxton, 2017).

Three Reasons Why Randomisation is Important

The purpose of the allocation concealment is to prevent the selection bias. Hence, it facilitates “enrolment of comparable participants” in each study group (Ruxton, 2017). The purpose of blinding or masking   is to prevent the ascertainment or biases related to performance or attrition. Blinding facilitates comparable concomitant care and evaluates participant in each study group (Pocock, 2013).

Intention to treat analysis or ITT means including all the patients in the analysis who were randomly allocated or enrolled in the group that is randomised. This process aims to avoid different misleading artifacts likely to occur in intervention research (Hollis & Campbell, 1999). In short this method estimates the effect of treatment as “offered or as assigned”.

In this process, inclusion occurs irrespective of the deviations that have the chance to occur after the process of randomisation such as withdrawal from the study or non-compliance. This analysis is preferred to any other type of analysis because it preserves the “prognosis balance” generated by original random treatment allocation. This analysis gives reliable estimate of effectiveness of true treatment by replicating things as they happen in real world (actual practice). This analysis prevents the bias in outcome and minimises type 1 statistical error. The estimates of the treatment effect are more conservative when compared to the PP analysis (Dossing et al., 2016). The nature of this analysis is simple when compared to other forms of analysis. Unlike other forms of analysis, ITT gives information on “potential effects of treatment policy” instead on potential effect of “specific treatment”. Moreover in this analysis validity for the study sample follows from the protection against “unmeasured confounding”. This is done without adjusting for non-adherence, which maximises the external validity (Johnston & Guyatt, 2016).  It is commonly preferred in comparison to other mode of analysis in randomised control trials because of its benefits on external validity (Hollis & Campbell, 1999).

Definition

Explanatory trial is a clinical trial used to test the possible beneficial effect of an intervention in an ideal situation (Sedgwick, 2014).

Pragmatic trial is a clinical trial designed to measure the effectiveness or extent of the positive effect of intervention in real clinical practice (Schulz, 1995).

Comparison of pragmatic and explanatory trials

Explanatory trials are conducted in large referral based health centres or in tertiary care. On the other hand, the pragmatic trials are conducted under real world conditions.  

Explanatory trials are used to assess if a drug or a therapy can work whereas the pragmatic trials tests if a therapy or drug does in fact work. Unlike the explanatory trials the pragmatic tests do not apply the stringent exclusion criteria. The former applies strict participant eligibility criteria. The eligibility is restricted to highly responsive, high risk, and highly compliant patients (Pocock, 2013). On the other hand, the pragmatic tests are free to participate by everyone with the condition of interest. In case of the explanatory trials the experimental intervention are inflexible with strict instructions for every element. In these trials both the practioner and the participants are usually blind. These trials prohibit the crossovers. In contrast, the pragmatic trials are used in routine health care so are highly flexible. In this trial no one is blind and permits crossovers. The comparison intervention in the explanatory trials often employs placebo with other conditions being same as experimental intervention. In pragmatic trials, there is a usual care for this condition (Sedgwick 2014).

Definition and Importance of Allocation Concealment

In explanatory trials the practioners and settings have high expertise whereas full range of practioners are present in the pragmatic trials with setting where a successful interventions would be applied. The compliance of the participants with the intervention is closely monitored in the explanatory trials and is a prerequisite for entry into study. High compliance is used for both the prophylactic strategies and the rescue strategies. On the other hand in pragmatic trial there is no measurement of compliance (Larson et al., 2016). In explanatory trials, the practioners’ adherence to the protocol is closely monitored and triggers various intervention whenever there is a deficient. There is no such action in the pragmatic trials (Roland & Torgerson, 1998).

The explanatory trials have frequent follow up intensity with extensive data collection. The blinded experts often determine the primary outcome which are surrogate outcomes. The primary analysis may try to justify excluding the non-responders. In pragmatic trials there is a usual follow up intensity and the primary outcome is determined in the routine course of healthcare. The primary analysis never deviates from the intention to treat all the patients in the trial (Sedgwick 2014).   

Usefulness

Pragmatic trials are more useful than the explanatory trials because the outcomes are directly relevant to the patients, health care practioners, funders and the communities. Since it does not use placebo like explanatory the effectiveness is estimated. These trial have conditions that mimic the routine clinical practice. Therefore, the results are more applicable to the “average patient” (Larson et al., 2016).

References 

Altman, D. G., & Schulz, K. F. (2001). Concealing treatment allocation in randomised trials. Bmj, 323(7310), 446-447.

Day, S. J., & Altman, D. G. (2000). Blinding in clinical trials and other studies. Bmj, 321(7259), 504.

Dossing, A., Tarp, S., Furst, D. E., Gluud, C., Wells, G. A., Beyene, J., ... & Christensen, R. (2016). Modified intention-to-treat analysis did not bias trial results. Journal of clinical epidemiology, 72, 66-74.

Gore, S. M. (1981). Assessing clinical trials--why randomise?. British medical journal (Clinical research ed.), 282(6280), 1958.

Hollis, S., & Campbell, F. (1999). What is meant by intention to treat analysis? Survey of published randomised controlled trials. Bmj, 319(7211), 670-674.

Johnston, B. C., & Guyatt, G. H. (2016). Best (but oft-forgotten) practices: intention-to-treat, treatment adherence, and missing participant outcome data in the nutrition literature. The American Journal of Clinical Nutrition, 104(5), 1197-1201.

Larson, E. B., Tachibana, C., Thompson, E., Coronado, G. D., DeBar, L., Dember, L. M., ... & Septimus, E. (2016, September). Trials without tribulations: Minimizing the burden of pragmatic research on healthcare systems. In Healthcare (Vol. 4, No. 3, pp. 138-141). Elsevier.

Paludan-Müller, A., Laursen, D. R. T., & Hróbjartsson, A. (2016). Mechanisms and direction of allocation bias in randomised clinical trials. BMC Medical Research Methodology, 16(1), 133.

Pocock, S. J. (2013). Clinical trials: a practical approach. John Wiley & Sons.

Roland, M., & Torgerson, D. J. (1998). Understanding controlled trials: What are pragmatic trials?. BMJ: British Medical Journal, 316(7127), 285.

Ruxton, G. D. (2017). Allocation concealment as a potentially useful aspect of randomised experiments. Behavioral Ecology and Sociobiology, 71(2), 31.

Schulz, K. F. (1995). Subverting randomization in controlled trials. Jama, 274(18), 1456-1458.

Sedgwick, P. (2013). Allocation concealment versus blinding in randomised controlled trials. Bmj, 347, f5518.

Sedgwick, P. (2014). Explanatory trials versus pragmatic trials. BMJ: British Medical Journal, 349.